With the development of video communication technologies and the improvement of network bandwidth, more and more users adopt video communication, such as network video playing, video chat, video conference and video-on-demand. Thus, life and work of societal public have been greatly enriched. Meanwhile, further development of the video communication technologies and network technologies are accelerated.
However, since the instability of network bandwidth and network transmission quality easily caused by complicated architecture and “best effort” characteristics of the Internet, video communication quality of current video communication system is difficult to be guaranteed under some circumstances, which has a serious influence on users' usage. One of primary bottlenecks affecting video communication quality currently is pause of video transmission caused by loss of network packets. FIG. 1 is a brief data flow diagram illustrating existed video communication technologies, in which video encoding end encodes video frames according to predefined rules and frame type, and then sends the encoded video frames to video decoding end through a network. The video decoding end decodes received video frames, and then plays the video.
However, according to currently widely applied technical solution, when network packets are lost in network congestion caused by a certain reason and video decoding end cannot restore decodable data with remaining data, it is necessary for the video encoding end to encode an Intra frame (I frame) with large data amount to be sent to the video decoding end, thus, not only bandwidth resources are wasted, but also existed network congestion is further aggravated.